interop

class pylibcudf.interop.ColumnMetadata(name: str = '', timezone: str = '', precision: int | None = None, children_meta: list[~pylibcudf._interop_helpers.ColumnMetadata] = <factory>)

Metadata associated with a column.

This is the Python representation of cudf::column_metadata.

Attributes

precision

pylibcudf.interop.from_dlpack(managed_tensor, Stream stream=None, DeviceMemoryResource mr=None) → Table

Convert a DLPack DLTensor into a cudf table.

For details, see from_dlpack()

Parameters:

managed_tensorPyCapsule

A 1D or 2D column-major (Fortran order) tensor.

streamStream | None

CUDA stream on which to perform the operation.

mrDeviceMemoryResource | None

Device memory resource used to allocate the returned table’s device memory.

Returns:

Table

Table with a copy of the tensor data.

pylibcudf.interop.to_dlpack(Table input, Stream stream=None, DeviceMemoryResource mr=None)

Convert a cudf table into a DLPack DLTensor.

For details, see to_dlpack()

Parameters:

inputTable

A 1D or 2D column-major (Fortran order) tensor.

streamStream | None

CUDA stream on which to perform the operation.

mrDeviceMemoryResource | None

Device memory resource used to allocate the returned DLPack tensor’s device memory.

Returns:

PyCapsule

1D or 2D DLPack tensor with a copy of the table data, or nullptr.